Transglutaminases (TGases) catalyze the formation of a crosslink between a donor amide group of a protein-bound glutamine residue and an acceptor e-NH2 of a protein-bound lysine residue. This crosslink is an isopeptide bond that cannot be cleaved in eukaryote organisms. The net result therefore is the formation of a permanent, stable, insoluble macromolecular protein complex. In the epidermis and other stratified squamous epithelia, at least three different TGase enzymes, TGases 1, 2 and 3, are expressed. They crosslink a variety of defined structural proteins to form the cornified cell envelope which is a principal component of epithelial barrier function. We are studying each of these enzymes in detail. Transglutaminase 1 The TGase 1 enzyme in cultured keratinocytes or foreskin epidermal cells is complex since it exists in multiple soluble and membrane-bound full-length as well as proteolytically-processed forms. The partitioning between the cytosol and membranes is controlled by differential acylation by myristate and palmitate of a cluster of cysteine residues located on a membrane anchorage amino-terminal segment which is unique to the TGase 1 enzyme. The various forms display wide variations in specific activities, but these are difficult to measure because the enzyme is inherently unstable and easily degraded by proteolysis. To address this problem, we have developed methods for its expression in the baculovirus system in which it is postsynthetically modified essentially the same ways as in keratinocytes, but as it is not proteolyzed it is more stable. In this way, we are able to produce microg amounts of enzyme which can be stored for use indefinitely. Previous work from this laboratory has shown that mutations in the TGM1 gene, encoding the TGase 1 enzyme, cause the autosomal recessive disorder lamellar ichthyosis. We have expressed in baculovirus some of the known mutations. Most result in a product of no or only very low activity. However, two others result in a product of >10-fold activity than the wildtype enzyme. Examination of their structures using the coordinates of the available structure of the related factor XIIIa TGase enzyme, we anticipate that these two mutations result in an unusually stabilized protein. In order to test this idea, we next expressed these same forms with an attached histidine tag at the carboxy terminus, and used them for transfection experiments into cultured keratinocytes. By specific immuno-precipitations with the his-tag antibody, we found that these mutant proteins were not proteolytically processed into highly active forms in vivo. These data suggest that lamellar ichthyosis disease may be caused by either insufficient enzyme activity, or an enzyme form that is not appropriately postsynthetically modified and cannot be utilized by the cell. Ongoing work is directed toward an understanding of the mechanism by which the TGase 1 enzyme is anchored to membranes and how this affects its substrate specificity toward certain known substrates including loricrin, small proline rich proteins, involucrin and various members of the plakin family including envoplakin and desmoplakin. In addition, available data suggest the 90 residue membrane anchorage segment controls TGase 1 activity. We will attempt to express this fragment in bacteria/baculovirus systems in order to study its structure and functional properties. Transglutaminase 2Our main focus of this enzyme is to obtain atomic resolution structural information by X-ray diffraction of crystals. To date, we have developed methods for the large scale preparation of the enzyme in baculovirus. Ongoing work will attempt to purify active forms of the enzyme in order to initiate crystallization trials. Transglutaminase 3The TGase 3 enzyme is expressed in many epithelial cell types, initially as an inactive pro-enzyme, that requires proteolytic activation by specific cleavage. In addition, data from this laboratory have shown that it is the preferred enzyme for crosslinking in vivo of several important substrates involved in barrier or other functions, including loricrin, small proline rich proteins, and trichohyalin. The proximal promoter region of the TGM 3 gene is located within the first 126 bp above the transcription start site, and consists of an Sp1 motif modulated by adjacent ets-like motifs. These are sufficient to confer epithelial-specific expression. This region also contains a calcium responsive element. In addition, we have found evidence for a single-stranded DNA binding protein that may serve as a negative controlling element. Studies on this promoter will continue. Activatable pro-TGase 3 enzyme has been expressed in large quantities in baculovirus. Preliminary work has demonstrated the formation of small crystals, from which we have generated structural information at the 3.5 A level using the Brookhaven synchronton X-ray facility. We are now developing procedures for the growth of larger crystals that may be suitable for complete structural ascertainment. We have found that labeling during synthesis in baculovirus with selenomethionine may enhance the X-ray signals and resolution. These studies will continue. Expression of TGases in non-epithelial tissues: roles in degenerative diseasesBy RT-PCR methods, we have found that TGases 1 and 3 are widely expressed in a variety of non-epithelial tissues, including in particular, various tissues within the brain, connective tissues, fibroblasts and muscle. Heretofore, these tissues were thought to express only the cytosolic TGase 2 enzyme. The expression of these enzymes has been confirmed by both immunoprecipitation reactions with specific antibodies, and by indirect immunofluorescence methods. Moreover, mRNA and enzyme levels of TGases 1 and 3 are upregulated in pathological conditions, including Alzheimer?s Disease (AD) and sporadic inclusion body myositis (SIBM). We found co-localization of the TGase enzyme antigens with the inclusion bodies/senile plaques of both diseases. In addition, we isolated insoluble proteins from tissue slices from these AD and SIBM which contained substantial amounts of isopeptide crosslink formed by TGases, suggesting a direct causative link. Also, in the case of SIBM, we performed sequencing analyses to demonstrate the presence of significant amounts of b-amyloid protein crosslinked to itself, myosin and desmin muscle proteins. Together, these data demonstrate that elevated levels of TGases 1 and/or 3 correlate with disease pathogenesis and contribute directly to the formation of insoluble crosslinked bodies that interfere with normal cellular function and thus degenerative disease. Depending on availability of adequate amounts of tissue, further work will be directed toward analysis of the forms of the TGase 1 and 3 enzymes and more detailed sequencing analyses. We anticipate that such studies may provide valuable new insights into both disease etiology as well as methods to control the degenerative diseases processes. Finally, we have found that when fibroblasts are irritated by low levels of a reagent such as SDS, they greatly elevate the expression of the TGase 1 enzyme, which begins to crosslink various cytoskeletal proteins including vimentin and actin. Preliminary data suggest this does not promote apoptosis. However, this system may serve as an experimental model of contact dermatitis.